A self-service system for multiple game players

Abstract

The present invention discloses a self-service system for multiple game players, and belongs to the field of computer technology. When a game user asks a question through a retrieval unit, the self-service system can automatically and proactively reply to the game user according to historical records. Compared with the manual processing method of customer service, the system can help reduce labor costs and improve reply efficiency. In addition, the present invention takes into account that during the game, many questions have the same meaning, but they correspond to questions with different answers at different times, and analyzes the relationship between different reply texts and reference standards such as game progress to obtain the association between different standard reply texts and reference standards such as game progress. When a game user asks a question later, the standard reply text that is most suitable for the game user is judged according to these reference standards, so that the game user can obtain the reply text that is closest to the actual need, which significantly improves the accuracy of the reply.

Description

Technical Field

[0001] The present invention belongs to the field of computer technology, and in particular, relates to a self-service system for multiple game players. Background Art

[0002] With the rapid development of the game industry at home and abroad, the market of the game industry has grown significantly, the number of game players has continued to rise, and the various needs of players have gradually increased. Relying solely on customer service calls to handle various needs of players is inefficient and costly. If there is a shortage of manpower, player needs have not been handled, and the user experience is poor, which will lead to a large number of users leaving. Maintaining players is an indispensable part of game operations, so a product is needed to maintain the various needs of players and better serve players. At present, many domestic games have developed player maintenance centers, but there are still the following problems: 1. Lack of intelligent reply function, all feedback needs to be manually processed by customer service, which has high labor costs and long processing time; 2. Lack of self-service inquiry and self-service billing functions, all can only be consulted by phone; In order to solve the above problems, a method is provided to accurately reply to and meet the questions of game players, and the present invention provides the following technical solutions. Summary of the invention

[0003] The purpose of the present invention is to provide a self-service system for multiple game players, so as to solve the problem that the responses to the game players' needs in the prior art mainly rely on manual processing and the accuracy of the intelligent responses is poor.

[0004] The purpose of the present invention can be achieved by the following technical solutions:

[0005] A self-service system for multiple game players, comprising:

[0006] A retrieval unit, through which a game user inputs retrieval information, and retrieves the standard reply text in the reply matching unit according to the retrieval information;

[0007] An intention recognition unit, used to obtain the question intention of each question text;

[0008] The answer matching unit is used to save the standard answer text and the answer text, and establish the correspondence between the standard answer text, the answer text and the question intention;

[0009] A keyword recognition unit, used for recognizing keywords in the reply text;

[0010] A game auxiliary unit for reading the reference standard of the game user;

[0011] The reference standard is a standard related to the time when the question text corresponding to the question intention is raised, and the reference standard includes time, game progress, etc.;

[0012] The controller is used to identify and obtain the standard reply text corresponding to each question intention, and is also used to determine the final reply text corresponding to the question intention, and send the final reply text corresponding to the determined question intention to the terminal device of the game user.

[0013] As a further solution of the present invention, the method for the controller to determine the final answer text corresponding to the question intention includes the following steps:

[0014] S1. Obtain the question text of the game user, obtain the intention of each question text through the intention recognition unit, and mark the intention as the question intention;

[0015] Get the answer text corresponding to each question text;

[0016] The question texts with the same question intention are considered as a question text group;

[0017] Get the answer text corresponding to each question text in the same question text group;

[0018] Divide the answer texts corresponding to each question text in the same question text group into one or more alienated groups;

[0019] The keyword recognition unit obtains the keywords of each reply text, the keyword similarity of the reply texts in the same alienated group is greater than or equal to a preset ratio θ, and the keyword similarity of any two reply texts between two alienated groups corresponding to the same question text group is less than θ;

[0020] The controller selects the best answer text as the standard answer text corresponding to the question intention according to the game user feedback;

[0021] When a question intention corresponds to two or more alienated groups, then the question intention corresponds to a corresponding number of standard answer texts;

[0022] S2, obtaining the approval standards corresponding to each heterogeneous group of each questioning intention, and obtaining the standard reference value R of each approval standard corresponding to the questioning intention;

[0023] S3. When the game user inputs question information through the retrieval unit, the question information input by the game user is analyzed by the intention recognition unit to obtain the question intention of the game user, and the corresponding standard answer text is matched in the answer matching unit according to the question intention of the game user;

[0024] When the game user's question intention corresponds to only one standard answer text, the standard answer text is used as the final answer text of the game user's question intention, and the final answer text is sent to the game user's terminal device;

[0025] When the game user's question intent corresponds to two or more standard answer texts:

[0026] First, obtain the approval standards corresponding to each differentiation group and the standard reference value Rj of each approval standard corresponding to each differentiation group;

[0027] Obtain the actual reference value Csj of each approved standard currently corresponding to the game user who asked the question through the game auxiliary unit;

[0028] According to the formula Calculate the correlation coefficient G between the standard answer text corresponding to each heterogeneous group and the game user who asked the question;

[0029] Where k is the number of approved standards corresponding to the corresponding alienation group;

[0030] The correlation coefficient G between the standard answer text corresponding to each alienated group and the game user who asked the question is calculated in sequence;

[0031] The correlation coefficients G are marked in ascending order as G1, G2, ..., Gx; where x is the number of alienated groups corresponding to the game user's question intention;

[0032] When (G2-G1) / G1≥β1 holds true, the standard reply text of the alienated group corresponding to G1 is considered to be the final reply text of the game user's question intention, and the final reply text is sent to the game user's terminal device; β1 is a preset proportional coefficient;

[0033] When (Gv-G1) / G1<β1 holds, the corresponding Gv and G1 are considered to be approximate items, all approximate items of G1 are obtained, and the standard reply texts of G1 and all approximate items of G1 are sent to the game user's terminal device in sequence as the final reply text of the game user's question intention; where 1≤v≤x.

[0034] As a further solution of the present invention, the value of θ is 85%.

[0035] As a further solution of the present invention, the method in which the controller selects the best reply text as the standard reply text corresponding to the question intention according to the game user feedback comprises the following steps:

[0036] For the reply texts in the same alienated group, obtain the time difference t1 between the time when each reply text was generated and the current time, the number of clicks d of each reply text, and the average score h of each reply text;

[0037] Then according to the formula Calculate the evaluation value H of each reply text;

[0038] The response text composition with the largest evaluation value H corresponds to the standard response text corresponding to the alienated group.

[0039] As a further solution of the present invention, a method for obtaining each approval standard corresponding to the question intention and the standard reference value R of each approval standard is as follows: for a question intention, mark it as a sample intention, obtain a question text group corresponding to the sample intention, and mark the question text in the question text group as a sample question text;

[0040] Obtain reference values ​​corresponding to each sample question text;

[0041] The reference value is the value corresponding to the reference standard, and the unit of the reference value varies according to the specific reference standard;

[0042] In an alienated group corresponding to the question intention, m reference values ​​corresponding to a reference standard corresponding to the alienated group are marked C1, C2, ..., Cn in sequence, where n is the number of sample question texts in the question text group corresponding to the sample intention;

[0043] According to the formula Calculate the confusion coefficient F of these n reference values;

[0044] When F≤F1, the corresponding reference standard is considered to be the approved standard of the corresponding alienation group, and Cp is used as the standard reference value R of the corresponding reference standard relative to the corresponding alienation group;

[0045] Wherein 1≤i≤n, Cp=(C1+C2+、…、+Cn) / n.

[0046] Beneficial effects of the present invention:

[0047] 1. The present invention can analyze the question texts raised by game users and the reply texts made by customer service. When game users ask questions through the search unit, the present invention can automatically and proactively reply to game users based on historical records. Compared with the manual processing method of customer service, it can help reduce labor costs and improve reply efficiency;

[0048] 2. The present invention takes into account that during the course of the game, many questions have the same meaning, but have different answers at different times. By analyzing the relationship between different reply texts and reference standards such as the game process, the association between different standard reply texts and reference standards such as the game process is obtained. When a game user asks a question later, the game user's current game process and other reference standards can be obtained, and the standard reply text that best suits the game user can be judged based on these reference standards, so that the game user can obtain the reply text that is closest to the actual need, which significantly improves the accuracy of the reply. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The present invention will be further described below in conjunction with the accompanying drawings.

[0050] Figure 1 It is a schematic diagram of the framework structure of a self-service system for multiple game players of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] A self-service system for multiple game players, such as Figure 1 As shown, including:

[0053] A retrieval unit, through which a game user inputs retrieval information, and retrieves the standard reply text in the reply matching unit according to the retrieval information;

[0054] The intention recognition unit analyzes the question text through the intention recognition algorithm to obtain the question intention of each question text;

[0055] The question text refers to the text of the question asked by the game user in the form of telephone or text;

[0056] The answer matching unit is used to save the standard answer text and the answer text, and establish the correspondence between the standard answer text, the answer text and the question intention;

[0057] The reply text refers to the text in which the customer service replies to questions raised by game users in the form of telephone or text;

[0058] A keyword recognition unit, used for recognizing keywords in the reply text;

[0059] A game auxiliary unit for reading the reference standard of the game user;

[0060] The controller is used to identify and obtain the standard answer text corresponding to each question intention, and is also used to determine the final answer text corresponding to the question intention, and send the determined final answer text corresponding to the question intention to the terminal device of the game user;

[0061] The steps include:

[0062] S1. Obtain the question text of the game user, analyze the question text through the intention recognition unit, obtain the intention of each question text, and mark the intention as the question intention;

[0063] Get the answer text corresponding to each question text;

[0064] The question texts with the same question intention are considered as a question text group;

[0065] Get the answer text corresponding to each question text in the same question text group;

[0066] Divide the answer texts corresponding to each question text in the same question text group into one or more alienated groups;

[0067] The keyword recognition unit is used to obtain the keywords of each reply text, the keyword similarity of the reply texts in the same alienated group is greater than or equal to a preset ratio θ, and the keyword similarity of any two reply texts between two alienated groups corresponding to the same question text group is less than θ; in one embodiment of the present invention, the value of θ is 85%;

[0068] The controller selects the best answer text as the standard answer text corresponding to the question intention according to the game user feedback;

[0069] When a question intention corresponds to two or more alienated groups, then the question intention corresponds to a corresponding number of standard answer texts, and each standard answer text corresponds to an alienated group;

[0070] After the standard reply text is selected, it can be re-edited by the staff to improve the quality of the reply text;

[0071] In one embodiment of the present invention, the method in which the controller selects the best answer text as the standard answer text corresponding to the question intention according to the game user feedback comprises the following steps:

[0072] For the reply texts in the same alienated group, obtain the time difference t1 between the time when each reply text was generated and the current time, the number of clicks d of each reply text, and the average score h of each reply text;

[0073] Then according to the formula Calculate the evaluation value H of each reply text;

[0074] The response text composition with the largest evaluation value H corresponds to the standard response text corresponding to the alienated group;

[0075] S2, obtaining the reference standards corresponding to each heterogeneous group of each questioning intention, and obtaining the standard reference value R of each reference standard corresponding to the questioning intention;

[0076] Specific:

[0077] For a question intention, mark it as a sample intention, obtain a question text group corresponding to the sample intention, and mark the question text in the question text group as a sample question text;

[0078] Obtain reference values ​​corresponding to each sample question text;

[0079] The reference value is the value corresponding to the reference standard, and the unit of the reference value varies according to the specific reference standard;

[0080] Taking an alienated group corresponding to the question intention as an example, m reference values ​​corresponding to a reference standard corresponding to the alienated group are marked C1, C2, ..., Cn in sequence, where n is the number of sample question texts in the question text group corresponding to the sample intention;

[0081] According to the formula Calculate the confusion coefficient F of these n reference values;

[0082] When F≤F1, the corresponding reference standard is considered to be the approved standard of the corresponding alienation group, and Cp is used as the standard reference value R of the corresponding reference standard relative to the corresponding alienation group;

[0083] Where 1≤i≤n, Cp=(C1+C2+,…,+Cn) / n;

[0084] The reference standard is a standard related to the time when the question text corresponding to the question intention is raised, and the reference standard includes time, game progress, etc.;

[0085] For ease of understanding, the following examples are given:

[0086] When a game user reaches a certain level in the game, he will encounter a specific question. At this time, a certain number of game users will ask the same question through the search unit at this progress, that is, the question is related to the progress of the game;

[0087] Here are some more examples:

[0088] During a certain period of time, due to the operation of the game itself, there are specific problems. During this period of time, a certain number of game users will frequently ask the same question through the search unit, that is, the question is related to the corresponding time period;

[0089] S3. When the game user inputs question information through the retrieval unit, the question information input by the game user is analyzed by the intention recognition unit to obtain the question intention of the game user, and the corresponding standard answer text is matched in the answer matching unit according to the question intention of the game user;

[0090] When the game user's question intention corresponds to only one standard answer text, the standard answer text is used as the final answer text of the game user's question intention, and the final answer text is sent to the game user's terminal device;

[0091] When the game user's question intention corresponds to two or more standard answer texts:

[0092] First, obtain the approval standards corresponding to each differentiation group and the standard reference value Rj of each approval standard corresponding to each differentiation group;

[0093] Obtain the actual reference value Csj of each approved standard currently corresponding to the game user who asked the question through the game auxiliary unit;

[0094] According to the formula Calculate the correlation coefficient G between the standard answer text corresponding to each heterogeneous group and the game user who asked the question;

[0095] Where k is the number of approved standards corresponding to the corresponding alienation group;

[0096] According to the above method, the correlation coefficient G between the standard reply text corresponding to each alienation group and the game user who asked the question is calculated in sequence;

[0097] The correlation coefficients G are marked in ascending order as G1, G2, ..., Gx;

[0098] Where x is the number of alienated groups corresponding to the game user’s question intention;

[0099] When (G2-G1) / G1≥β1 holds true, the standard reply text of the alienated group corresponding to G1 is considered to be the final reply text of the game user's question intention, and the final reply text is sent to the game user's terminal device;

[0100] Where β1 is the preset proportional coefficient;

[0101] When (Gv-G1) / G1<β1 holds, it is considered that the corresponding Gv and G1 are approximate items, all approximate items of G1 are obtained, and the standard reply texts of G1 and all approximate items of G1 are sent to the terminal device of the game user as the final reply text of the game user's question intention in sequence;

[0102] Where 1≤v≤x;

[0103] The present invention takes into account that during the game, many questions have the same meaning, but at different times they correspond to questions with different answers. Therefore, the prior art method of providing intelligent answers by presetting answers and based on the interpretation of questions will have a high error rate at this time, and cannot meet the intelligent answer needs of game users. The present application analyzes the relationship between different answer texts and reference standards such as game progress, obtains the association between different standard answer texts and reference standards such as game progress, and when a game user asks a question later, the game user's current game progress and other reference standards can be obtained, and the standard answer text that is most suitable for the game user can be judged based on these reference standards, so that the game user can obtain the answer text that is closest to the actual need, which significantly improves the accuracy of the answer.

[0104] In the description of the specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0105] The above contents are merely examples and explanations of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the invention or exceed the scope defined by the claims, they shall all fall within the protection scope of the present invention.

Claims

1. A self-service system for multiple game players, characterized in that: include: A retrieval unit, through which a game user inputs retrieval information, and retrieves the standard reply text in the reply matching unit according to the retrieval information; An intention recognition unit, used to obtain the question intention of each question text; The answer matching unit is used to save the standard answer text and the answer text, and establish the correspondence between the standard answer text, the answer text and the question intention; A keyword recognition unit, used for recognizing keywords in the reply text; A game auxiliary unit for reading the reference standard of the game user; The reference standard is a standard related to the time when the question text corresponding to the question intention is raised, and the reference standard includes time and game progress; The controller is used to identify and obtain the standard answer text corresponding to each question intention, and is also used to determine the final answer text corresponding to the question intention, and send the determined final answer text corresponding to the question intention to the terminal device of the game user; The method for the controller to determine the final answer text corresponding to the question intention includes the following steps: S1. Obtain the question text of the game user, obtain the intention of each question text through the intention recognition unit, and mark the intention as the question intention; Get the answer text corresponding to each question text; The question texts with the same question intention are regarded as a question text group; Get the answer text corresponding to each question text in the same question text group; Divide the answer texts corresponding to each question text in the same question text group into one or more alienated groups; The keyword recognition unit obtains the keywords of each reply text, the keyword similarity of the reply texts in the same alienated group is greater than or equal to a preset ratio θ, and the keyword similarity of any two reply texts between two alienated groups corresponding to the same question text group is less than θ; The controller selects the best answer text as the standard answer text corresponding to the question intention according to the game user feedback; When a question intention corresponds to two or more alienated groups, then the question intention corresponds to a corresponding number of standard answer texts; S2, obtaining the approval standards corresponding to each heterogeneous group of each questioning intention, and obtaining the standard reference value R of each approval standard corresponding to the questioning intention; S3. When the game user inputs question information through the retrieval unit, the question information input by the game user is analyzed by the intention recognition unit to obtain the question intention of the game user, and the corresponding standard answer text is matched in the answer matching unit according to the question intention of the game user; When the game user's question intention corresponds to only one standard answer text, the standard answer text is used as the final answer text of the game user's question intention, and the final answer text is sent to the game user's terminal device; When the game user's question intent corresponds to two or more standard answer texts: First, obtain the approval standards corresponding to each differentiation group and the standard reference value Rj of each approval standard corresponding to each differentiation group; Obtain the actual reference value Csj of each approved standard currently corresponding to the game user who asked the question through the game auxiliary unit; According to the formula Calculate the correlation coefficient G between the standard answer text corresponding to each heterogeneous group and the game user who asked the question; Where k is the number of approved standards corresponding to the corresponding alienation group; The correlation coefficient G between the standard answer text corresponding to each alienated group and the game user who asked the question is calculated in sequence; The correlation coefficients G are marked in ascending order as G1, G2, ..., Gx; where x is the number of alienated groups corresponding to the game user's question intention; When (G2-G1) / G1≥β1 holds true, the standard reply text of the alienated group corresponding to G1 is considered to be the final reply text of the game user's question intention, and the final reply text is sent to the game user's terminal device; β1 is a preset proportional coefficient; When (Gv-G1) / G1<β1 holds, the corresponding Gv and G1 are considered to be approximate items, all approximate items of G1 are obtained, and the standard reply texts of G1 and all approximate items of G1 are sent to the game user's terminal device in sequence as the final reply text of the game user's question intention; where 1≤v≤x.

2. A self-service system for multiple game players according to claim 1, characterized in that: The value of θ is 85%.

3. A self-service system for multiple game players according to claim 1, characterized in that: The method for the controller to select the best answer text as the standard answer text corresponding to the question intention according to the game user feedback includes the following steps: For the reply texts in the same alienated group, obtain the time difference t1 between the time when each reply text was generated and the current time, the number of clicks d of each reply text, and the average score h of each reply text; Then according to the formula Calculate the evaluation value H of each reply text; The response text composition with the largest evaluation value H corresponds to the standard response text corresponding to the alienated group.

4. A self-service system for multiple game players according to claim 1, characterized in that: The method for obtaining the approval standards corresponding to the question intention and the standard reference value R of each approval standard is as follows: for a question intention, mark it as a sample intention, obtain a question text group corresponding to the sample intention, and mark the question text in the question text group as a sample question text; Obtain reference values ​​corresponding to each sample question text; The reference value is the value corresponding to the reference standard, and the unit of the reference value varies according to the specific reference standard; In an alienated group corresponding to the question intention, m reference values ​​corresponding to a reference standard corresponding to the alienated group are marked C1, C2, ..., Cn in sequence, where n is the number of sample question texts in the question text group corresponding to the sample intention; According to the formula Calculate the confusion coefficient F of these n reference values; When F≤F1, the corresponding reference standard is considered to be the approved standard of the corresponding alienation group, and Cp is used as the standard reference value R of the corresponding reference standard relative to the corresponding alienation group; Wherein 1≤i≤n, Cp=(C1+C2+、…、+Cn) / n.

Images